This research aims to further understanding of the social epidemiology of emotional distress and well-being. In particular, sex differences in distress probably result from the different roles which women and men play and the fact that both demands and rewards of various roles depend on the sex of the incumbent. The three principal roles to be examined are marital, parental, and employment. Although the effects of these roles on distress have been investigated in varying detail, there is considerable confusion about the processes involved. Many differences are of marginal statistical significance, and researchers often report contradictory findings. These issues can be resolved best by an analysis of multiple data sets with multiple measures of both positive and negative affective states. Twelve data sets will be employed. The first step in the analysis will be to refine the internal consistency and criterion validity of the measures of distress. The estimation and testing of Rasch models appears promising. Once acceptable measures are developed, the various roles will be examined. Sex differences in distress are largest among the married but exist for all statuses but the widowed. Differences between married men and women in their experience of marriage will be examined in detail. Several surveys contain information on the division of household labor, the importance of marriage, how restrictive marriage is, and general marital values. These will be examined to determine whether they modify or explain the relationship between sex and distress among the married. The effects of the parental role are less well understood. Preliminary analysis will attempt to pinpoint the effects of the presence, number, ages, and clustering of children on fathers and mothers. Then the division of childcare responsibilities, attitudes about children, and related factors will be examined to see if they modify or explain the observed relationships. The last role to be examined is employment. Of particular interest are the effects of wives employment on their own and their husbands' mental health. There is some agreement that the negative effect of wives' employment on their husbands depends on the husband holding traditional sex role attitudes. Whether the beneficial impact of paid employment on the women depends on their attitudes is less clear and will be investigated. Finally, the combined nature of the effects will be investigated. Whether the effects are additive or multiplicative is of substantial theoretical importance.